File size: 2,304 Bytes
0fadc57 f4a9b3a 0fadc57 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 |
---
license: cc-by-2.0
datasets:
- CreativeLang/vua20_metaphor
language:
- en
---
# Metaphor_Detection_Roberta_Seq
## Description
- **Paper:** [FrameBERT: Conceptual Metaphor Detection with Frame Embedding Learning](https://aclanthology.org/2023.eacl-main.114.pdf)
## Model Summary
Creative Language Toolkit (CLTK) Metadata
- CL Type: Metaphor
- Task Type: detection
- Size: roberta-base (500MB)
- Created time: 2022
This model is a easy to use metaphor detection baseline realised with `roberta-base` fine-tuned on [CreativeLang/vua20_metaphor](https://huggingface.co/datasets/CreativeLang/vua20_metaphor) dataset.
To use this model, please use the `inference.py` in the [FrameBERT repo](https://github.com/liyucheng09/MetaphorFrame).
Just run:
```
python inference.py CreativeLang/metaphor_detection_roberta_seq
```
Check out `inference.py` to learn how to apply the model on your own data.
For the details of this model and the dataset used, we refer you to the release [paper](https://aclanthology.org/2023.eacl-main.114.pdf).
## Metrics
| Metric | Value |
|----------------------------------|--------------------------|
| eval_loss | 0.2656 |
| eval_accuracy_score | 0.9142 |
| eval_precision | 0.9142 |
| eval_recall | 0.9142 |
| eval_f1 | 0.9142 |
| eval_f1_macro | 0.7315 |
| eval_runtime | 8.9802 |
| eval_samples_per_second | 411.7960 |
| eval_steps_per_second | 51.5580 |
| epoch | 3.0000 |
### Citation Information
If you find this dataset helpful, please cite:
```
@article{Li2023FrameBERTCM,
title={FrameBERT: Conceptual Metaphor Detection with Frame Embedding Learning},
author={Yucheng Li and Shunyu Wang and Chenghua Lin and Frank Guerin and Lo{\"i}c Barrault},
journal={ArXiv},
year={2023},
volume={abs/2302.04834}
}
```
### Contributions
If you have any queries, please open an issue or direct your queries to [mail](mailto:yucheng.li@surrey.ac.uk). |